WebArena provides a realistic multi-domain web environment and benchmark where state-of-the-art LLM agents achieve 14.41% end-to-end task success compared to 78.24% for humans.
These situations typically indicate a high likelihood of execution failure and hence warrant early termination
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
citation-role summary
background 1
citation-polarity summary
fields
cs.AI 1years
2023 1verdicts
ACCEPT 1roles
background 1polarities
unclear 1representative citing papers
citing papers explorer
-
WebArena: A Realistic Web Environment for Building Autonomous Agents
WebArena provides a realistic multi-domain web environment and benchmark where state-of-the-art LLM agents achieve 14.41% end-to-end task success compared to 78.24% for humans.